I expect there will be as unique answers as there are architects; each of us seeing our world in just a bit different way from another. After reading the pages of threads with so much great content, its comforting (at least to me) that we are such passionate folk.
Back in graduate school so many years ago, there was a very smart fellow who wrote about software implementation. In simplified terms, he said you start with the essential functional criteria (as we do architecture) and work our way from piloting the software to the eventual supporting hardware, detailed components, and then training commitments. I still use this method today. Hopefully, in reading through these entries, you find the functionality you seek.
As an example, mine was:
1. Design visualization / strong modeling and open data platform vs a closed program - where the software development team navigates all manner of innovation trends. This was so I could explore more architecture as they explore hardware and software technology.
2. Design element tools over detailing tools. For me, a program I could use to design fluidly without resorting to yet another program, and remembering we are creating qualitative, buildable design 'intent' models; not shop drawings.
3. If not an Autodesk platform, then with quality import export to ACAD. (Much US government files and Record Drawings were mandated to be in ACAD format; may still do?) This allows me develop internal concepts in my 'platform of comfort' and pass them to consultants whom many have ADSK products.
4. Ability to connect to external databases (e.g., ODBC) for extended datasets and analysis.
5. MEP extension. (While I would not be designing the MEP systems, having the tools are of importance to me for the many benefits of building modeling gives.)
6. Total Cost of Ownership. (Initial license/s purchase + essential add-ons + hardware + training + annual upgrades (now subscription).
After working with a number of object-based and layer-based CAD programs in prior years, I chose ArchiCAD, 'by a length'. Every few years, I return to the questions above, albeit with some updates and emphasis. You may have other criteria:
- A client visualization tool (for example, BIMx vs BIM360).
- Production speed?
- Data-centric focus? Ease with IFC?
- Rendering tools are critical for some - internal engines vs export, compatibility.
- Algorithmic design is always emerging (Rhino? I'm thinking Paramo will be exciting.)
While your immediate firm platform is important, a personal criteria may be just as important. Past is not future, and we are in a huge digital evolution with more employment churn and international firm and 'gig' work on the rise. With a criteria in hand, you may be able to better your choice of a good ship.
“The best thing about the future is that it comes one day at a time.” - Abraham Lincoln
AC27 USA on 16” 2019 MBP (2.4GHz i9 8-Core, 32GB DDR4, AMD Radeon Pro 5500M 8G GDDR5, 500GB SSD, T3s, Trackpad use) running Sonoma OS + extended w/ (2) 32" ASUS ProArt PAU32C (4K) Monitors